Abstract
Objectives. We conducted a case–control study in the Jackson, Mississippi, area to identify factors associated with HIV infection among young African American men who have sex with men (MSM).
Methods. During February to April 2008, we used surveillance records to identify young (16–25 years old) African American MSM diagnosed with HIV between 2006 and 2008 (case participants) and recruited young African American MSM who did not have HIV (controls). Logistic regression analysis was used to assess factors associated with HIV infection.
Results. In a multivariable analysis of 25 case participants and 85 controls, having older male partners (adjusted odds ratio [OR] = 5.5; 95% confidence interval [CI] = 1.8, 17.3), engaging in unprotected anal intercourse with casual male partners (adjusted OR = 6.3; 95% CI = 1.8, 22.3), and being likely to give in to a partner who wanted to have unprotected sex (adjusted OR = 5.0; 95% CI = 1.2, 20.6) were associated with HIV infection.
Conclusions. Given the high prevalence of risk behaviors among the young African American MSM in our study, HIV prevention efforts must begin before or during early adolescence and need to focus on improving negotiation and communication regarding sex.
African American men who have sex with men (MSM) are at high risk of HIV infection.1,2 In the United States, half of all prevalent and incident HIV infections occur among MSM,3,4 and Black MSM account for a disproportionate number of new HIV infections among MSM.5 Data from the National HIV Behavioral Surveillance System show that the HIV prevalence among Black MSM in 21 US cities in 2008 was 28% and that 59% of Black MSM with HIV surveyed in these cities were unaware of their infection.6 HIV prevalence was also high among young Black MSM, who had a prevalence of 17% by the age of 18 to 24 years.6
Furthermore, the number of new HIV diagnoses is increasing among MSM, particularly young Black MSM. From 2001 to 2006, the number of HIV/AIDS cases among Black MSM aged 13 to 24 years in 33 states with long-term, confidential, name-based HIV reporting increased by 93%.7
In the fall of 2007, clinicians at a sexually transmitted disease (STD) clinic in Jackson, Mississippi, noted that diagnoses of HIV infection were increasing among young African American MSM. A subsequent review of surveillance data revealed a 38% rise in newly diagnosed HIV infections among African American MSM aged 16 to 25 years in the Jackson area during 2006–2007 relative to 2004–2005.
Although many studies have assessed the prevalence of HIV risk behaviors among young African American MSM,8,9 few have focused on young African American MSM in Mississippi. Between February and April 2008, the Centers for Disease Control and Prevention and the Mississippi State Department of Health conducted an investigation that included, among other components, a case–control study designed to identify demographic characteristics and behavioral factors associated with HIV infection among young African American MSM in the Jackson area.10
METHODS
In this unmatched case–control study, we defined case participants as young African American men (aged 16–25 years at the time of HIV diagnosis) who were diagnosed and reported with HIV infection during January 2006 to April 2008, lived in or were diagnosed in the Jackson area (Hinds, Madison, and Rankin counties), and reported having had anal sex with a man during the 12 months before their HIV diagnosis. We identified individuals using Mississippi HIV/AIDS Reporting System data on race, gender, age, date of diagnosis, and county of residence and diagnosis and considered them potential case participants pending confirmation of behavioral eligibility. We attempted to recruit all potential case participants by telephone, by mail, or in person. Potential case participants completed the entire self-administered survey (which was confidential and did not include names or other personally identifiable information) on a laptop or handheld computer. Those who did not report male–male sexual activity were excluded from the analysis.
We defined controls as young African American men aged 16 to 25 years who lived in the 3-county Jackson area, reported anal sex with a man during the preceding 12 months, and reported having had a negative HIV test during the 6 months before their interview. Controls were recruited from multiple venues, including an STD clinic, bars and clubs, college campuses, and shopping malls. At the STD clinic, the charts of all incoming patients were reviewed to determine age, race, and history of HIV infection. Young African American men 16 to 25 years of age who were not known to have HIV were approached, informed that a health survey was being conducted, and invited to complete the self-administered screening survey to determine their eligibility for participation.
At other venues, interviewers approached individuals who appeared to be African American men in the eligible age range, informed them about the survey, and invited them to complete the screening survey to determine eligibility. The screening survey collected information on age, race/ethnicity, county of residence, gender, sexual behavior, and HIV testing history and included questions about tobacco, alcohol, and drug use to minimize stigma associated with the survey by making it more difficult for those deemed not eligible to determine the eligibility criteria. Those who were determined to be eligible and consented to participate completed the full survey at that time. Individuals who did not report a negative HIV test in the preceding 6 months but met all other eligibility criteria were offered the opportunity to undergo an HIV test at the STD clinic and complete the full survey at that time. Those who tested positive were offered the opportunity to be surveyed as case participants.
Participation was voluntary, and case participants and controls who were eligible, provided consent, and completed the survey received a $25 gift card. Strict security and confidentiality measures consistent with those used for HIV surveillance data were used in maintaining all files containing survey data.
Survey Instrument and Measures
The survey was self-administered on a computer or handheld device and required approximately 30 minutes to complete. Unless otherwise noted, the recall periods were the 12-month period before the first positive HIV test for case participants and the 12-month period before the interview for controls.
Participants answered questions about sociodemographic characteristics, incarceration, sexual identity, substance use, and sexual behaviors. Questions were derived from a variety of sources, including the National HIV Behavioral Surveillance System instrument and survey instruments from prior outbreak investigations. Relationship status choices were as follows: single (never married), separated, divorced, in a long-term relationship with a man, in a long-term relationship with a woman, married to a man, and married to a woman.
All questions about sexual behaviors referred only to vaginal or anal sex. Participants were asked the number of male and female sex partners they had had during their lifetimes and during the recall period and the number of male partners they considered to be main partners (“partners that you have an emotional bond with and with whom you have regular sex, such as boyfriend, girlfriend, spouse, significant other, or life partner”) or casual partners (“those people you have sex with every now and then and one-night stands”). They were also asked about the age group (younger than 18 years, 18–25 years, 26–30 years, 31–40 years, older than 40 years) and race/ethnicity of their partners, condom use, concurrent relationships, exchange sex, and where they met partners during the recall period.
Respondents were considered to have had concurrent sexual relationships if they reported having 2 sex partners during the same week or having sex with one person while sexually involved with someone else. They were considered to have engaged in exchange sex if they reported giving or receiving something in exchange for sex, such as money, food, a place to stay, drugs or alcohol, or transportation. All respondents were asked to rate their agreement, on a 5-point scale ranging from strongly disagree to strongly agree, with the statement “If a partner wanted to have unprotected sex, I would probably give in.”
Data Analysis
The case–control analysis was limited to men who reported having had anal sex with a man during the 1-year recall period. In addition, we excluded from this analysis participants for whom data were missing for the key variables of unprotected anal intercourse, age of sex partners, and discussing HIV status with sex partners.
We used SAS version 9.2 (SAS Institute Inc, Cary, NC) in conducting all of the statistical analyses. Because of small sample sizes, we used the Fisher exact test to test for differences between case participants and controls with respect to categorical variables. We used the Wilcoxon–Mann–Whitney nonparametric test to compare the medians and distributions of continuous variables. Statistical significance was evaluated at P < .05.
We used multivariable logistic regression to identify factors independently associated with HIV infection. Because involvement in a long-term relationship with a man may affect condom use behaviors, we controlled for whether participants reported being in a long-term relationship with or married to a man. Sexual behaviors that were associated with HIV infection in the bivariate analysis at the P ≤ .1 level were eligible for inclusion in the multivariable analysis. Given the small number of observations in our sample, we used backward elimination to develop a parsimonious model including variables significant at P < .05. We estimated adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for HIV infection. We also assessed possible 2-way interactions between involvement in a long-term relationship, unprotected anal intercourse, and willingness to give in to a partner who wanted to have unprotected sex.
RESULTS
Between January 2006 and April 2008, 86 HIV infections among individuals who met the study criteria as potential case participants were reported to the Mississippi State Department of Health. Forty of these men (47%) were interviewed. Of the 46 not interviewed, 31 could not be contacted, 3 had moved out of the area, 1 was deceased, 1 declined to participate, 1 did not attend the scheduled interview, and 9 had no reason recorded. Those not interviewed were not significantly different from those interviewed with respect to age, year of diagnosis, risk category reported in the Mississippi HIV/AIDS Reporting System, or residency or HIV diagnosis in the Jackson area versus elsewhere in Mississippi. Of the 40 potential case participants who were interviewed, 30 (75%) were included in the case–control study. Those excluded did not report male–male sex (n = 9) or identified as transgender (n = 1).
Of the 936 potential controls who completed the screening survey, 95 (10%) were eligible and included in the case–control study. The most common reason for ineligibility was not reporting anal sex with a man in the preceding 12 months (88%).
We excluded from our analysis 5 case participants and 10 controls who did not answer all of the questions regarding key risk behaviors. Excluded case participants and controls were younger than those included in the analysis (median age = 19 vs 21 years; P = .01). There were no differences with respect to education, college enrollment, or residence in the Jackson area.
Case participants included in our analysis received their HIV diagnoses at the STD clinic (n = 16), at another clinic (n = 2), at a hospital (n = 2), at a blood donation center (n = 2), at a college testing program (n = 2), or in jail or prison (n = 1). Controls were recruited at bars or clubs (n = 30), college campuses (n = 21), an STD clinic (n = 19), shopping malls (n = 12), a social event for the African American gay community (n = 2), and a college testing program (n = 1).
Case participants were older than controls (median age = 22 vs 20 years; P = .004) and more likely to report being in a long-term relationship with or married to a man (Table 1). The self-reported sexual identity of case participants and controls did not differ significantly; most identified as gay or homosexual.
TABLE 1.
Selected Characteristics of Young African American Men Who Have Sex With Men: Jackson, MS, Area, 2006–2008
| Participants With HIV (n = 25), No. (%) | Participants Without HIV (n = 85), No. (%) | P | |
| Education levela | .5 | ||
| Less than high school | 1 (4) | 3 (4) | |
| High school or equivalent | 6 (24) | 30 (35) | |
| Some college or technical school or more | 18 (72) | 52 (61) | |
| Enrolled in college during recall periodb | 16 (64) | 70 (82) | .06 |
| Relationship statusa | <.001 | ||
| Single, never married | 15 (60) | 77 (91) | |
| Long-term relationship with or married to woman | 1 (4) | 0 (0) | |
| Long-term relationship with or married to man | 9 (36) | 7 (8) | |
| Divorced | 0 (0) | 1 (1) | |
| History of incarceration during recall periodb | 4 (16) | 9 (11) | .5 |
| Sexual identity | .3 | ||
| Gay/homosexual | 19 (76) | 52 (61) | |
| Bisexual | 3 (12) | 23 (27) | |
| Straight/heterosexual | 2 (8) | 4 (5) | |
| Other | 1 (4) | 6 (7) |
Note. P values were determined by the Fisher exact test. Numbers might not sum to column totals because of missing, “don't know,” or refused responses. Percentages might not sum to 100 because of missing, “don't know,” or refused responses or because of rounding.
At the time of the first positive HIV test (participants with HIV) or the interview (participants without HIV).
The recall periods were the 12 months before the first positive HIV test for participants with HIV and the preceding 12 months for participants without HIV.
Similar percentages of case participants and controls reported alcohol use (Table 2) during the recall period, and almost all who used alcohol reported doing so before or during sex. More than one third of case participants and controls reported using marijuana; few respondents reported using any other noninjection drugs. No participants reported injection drug use during the recall period.
TABLE 2.
Drug and Alcohol Use Among Young African American Men Who Have Sex With Men: Jackson, MS, Area, 2006–2008
| Participants With HIV (n = 25), No. (%) | Participants Without HIV (n = 85), No. (%) | P | |
| Alcohol use | 15 (60) | 62 (73) | .2 |
| Alcohol use before or during sex | 14 (93) | 48 (77) | .4 |
| Use of any noninjection drug | 10 (40) | 40 (47) | .6 |
| Use of any noninjection drug before or during sex | 9 (90) | 24 (60) | .1 |
| Marijuana use | 10 (40) | 31 (36) | .8 |
| Marijuana use before or during sex | 7 (70) | 17 (55) | .5 |
| Use of any other noninjection drug | 1 (4) | 14 (16) | .2 |
| Ecstasy use | 1 (100) | 11 (79) | .3 |
| Methamphetamine use | 0 (0) | 1 (7) | >.99 |
| Use of any other noninjection drug before or during sex | 1 (100) | 7 (50) | >.99 |
| Injection drug use | 0 (0) | 0 (0) | … |
Note. P values were determined by the Fisher exact test. The recall periods were the 12 months before the first positive HIV test for participants with HIV and the preceding 12 months for participants without HIV.
Drug and alcohol use were not significantly associated with HIV infection in the bivariate analysis (Table 2). In addition, median age at first sex with a woman, median age at first sex with a man, lifetime number of male sex partners, and number of male partners, main male partners, casual male partners, and female partners during the recall period were not associated with HIV infection (Table 3). Lastly, reports of male partners who were not African American, anonymous partners, exchange sex, concurrent sexual relationships, and meeting male sex partners on the Internet, at gay bars or clubs, or at nongay bars or clubs during the recall period were not associated with HIV infection.
TABLE 3.
Results of Bivariate Analysis of Potential Associations Between Sexual Behaviors and HIV Infection Among Young African American Men Who Have Sex With Men: Jackson, MS, Area, 2006–2008
| Participants With HIV (n = 25), Median (Range) or No. (%) | Participants Without HIV (n = 85), Median (Range) or No. (%) | P | Unadjusted Odds Ratio(95% Confidence Interval) | |
| Age, y, at first sex with a womana | 15.5 (12–22) | 15 (12–20) | .8 | … |
| Age, y, at first sex with a man | 17 (9–24) | 17 (2–23) | .7 | … |
| No. of male sex partners during lifetime | ||||
| 1 | 1 (4) | 4 (5) | >.99 | 0.8 (0.01, 10.4) |
| 2–4 | 4 (16) | 17 (20) | >.99 | 0.8 (0.2, 3.6) |
| 5–10 | 10 (40) | 32 (38) | >.99 | 1.1 (0.3, 3.6) |
| > 10 (Ref) | 8 (32) | 27 (32) | … | 1.0 |
| Lifetime no. of male sex partners | 8 (1–59) | 8 (1–42) | .7 | … |
| No. of main male sex partners during recall period | ||||
| 0 | 1 (4) | 11 (13) | .4 | 0.3 (0.01, 2.7) |
| 1 | 13 (52) | 31 (37) | .6 | 1.4 (0.5, 4.2) |
| 2–4 (Ref) | 10 (40) | 34 (40) | … | 1.0 |
| ≥ 5 | 1 (4) | 9 (11) | .7 | 0.4 (0.01, 3.4) |
| No. of casual male sex partners during recall period | ||||
| 0 | 9 (36) | 32 (38) | .4 | 0.6 (0.2, 2.0) |
| 1 | 3 (12) | 17 (20) | .3 | 0.4 (0.1, 1.9) |
| 2–4 (Ref) | 10 (40) | 22 (26) | … | 1.0 |
| ≥ 5 | 3 (12) | 14 (16) | .5 | 0.5 (0.1, 2.3) |
| No. of female sex partners during recall period | 2 (8) | 18 (21) | .2 | 0.3 (0.03, 1.5) |
| Unprotected vaginal intercourse with a female partner during recall period | 0 (0) | 5 (6) | >.99 | … |
|
Characteristics of and behaviors with sex partners | ||||
| Male sex partners older than 25 y during recall period | 15 (60) | 18 (21) | <.001 | 5.6 (1.9, 16.3) |
| Non–African American male sex partners during recall period | 4 (16) | 11 (13) | .7 | 1.3 (0.3, 4.9) |
| Main male sex partners during recall period | 24 (96) | 74 (87) | .3 | 3.6 (0.5, 158.7) |
| Unprotected anal intercourse with main male sex partners during recall period | 18 (75) | 35 (47) | .02 | 3.3 (1.1, 11.4) |
| Did not ask most recent main male sex partner his HIV statusb | 9 (38) | 11 (15) | .04 | 3.4 (1.04, 11.0) |
| Casual male sex partners during recall period | 16 (64) | 53 (62) | .9 | 0.9 (0.3, 2.6) |
| Unprotected anal intercourse with casual male sex partners during recall period | 12 (75) | 8 (15) | <.001 | 16.9 (3.7, 85.5) |
| Did not ask most recent casual male sex partner his HIV statusb | 6 (38) | 9 (17) | .1 | 2.9 (0.7, 11.8) |
| Anonymous male sex partners during recall period | 4 (16) | 10 (12) | .7 | 1.4 (0.3, 5.6) |
| Exchange sex during recall periodc | 4 (16) | 15 (18) | >.99 | 0.9 (0.2, 3.2) |
| Concurrent sexual relationships during recall periodd | 14 (56) | 48 (56) | >.99 | 1.0 (0.4, 2.7) |
| Met male sex partners on the Internet during recall period | 15 (60) | 49 (58) | >.99 | 1.1 (0.4, 3.1) |
| Met male sex partners at gay bars/clubs during recall period | 15 (60) | 39 (46) | .3 | 1.8 (0.7, 4.9) |
| Met male sex partners at nongay bars/clubs during recall period | 3 (12) | 20 (24) | .3 | 0.4 (0.1, 1.7) |
| Would probably give in if a partner wanted to have unprotected sexb | 8 (32) | 6 (7) | .003 | 6.2 (1.6, 24.3) |
Note. P values were determined by the Fisher exact test for categorical variables and the Wilcoxon—Mann—Whitney test for continuous variables. Numbers might not sum to column totals because of missing, “don't know,” or refused responses. Percentages might not sum to 100 because of missing, “don't know,” or refused responses or because of rounding. The recall periods were the 12 months before the first positive HIV test for participants with HIV and the preceding 12 months for participants without HIV. Main partners were defined as “partners that you have an emotional bond with and with whom you have regular sex, such as boyfriend, girlfriend, spouse, significant other, or life partner”; casual partners were defined as “those people you have sex with every now and then and one-night stands.”
Among the 16 participants with HIV and the 48 participants without HIV reporting any lifetime female sex partners.
At the time of the first positive HIV test among participants with HIV or the time of the interview among participants without HIV.
Included giving or receiving any of the following in exchange for sex: money, food, a place to stay, transportation, a promise not to hurt the participant or loved ones, drugs, or alcohol.
Respondents were considered to have had concurrent sexual relationships if they reported having 2 sex partners in the same week or having sex with one person while sexually involved with someone else.
Reporting male sex partners older than 25 years, reporting unprotected anal intercourse with main male sex partners, not asking one's most recent main male sex partner his HIV status, reporting unprotected anal intercourse with casual male partners, and agreeing or strongly agreeing that one would probably give in if a partner wanted to have unprotected sex were significantly associated with increased odds of HIV infection in the bivariate analysis (Table 3). Among case participants and controls who reported engaging in unprotected anal intercourse with their most recent main or casual partner, 54% reported not using a condom because “we got caught up in the moment.”
In the multivariable logistic regression controlling for whether respondents reported being in a long-term relationship with or married to a man, factors associated with HIV infection included having a male sex partner older than 25 years (adjusted OR = 5.5; 95% CI = 1.8, 17.3), having unprotected anal intercourse with a casual male sex partner (adjusted OR = 6.3; 95% CI = 1.8, 22.3), and agreeing that one would probably give in if a partner wanted to have unprotected sex (adjusted OR = 5.0; 95% CI = 1.2, 20.6) (Table 4).
TABLE 4.
Results of Multivariate Analysis of Factors Associated With HIV Infection Among Young African American Men Who Have Sex With Men: Jackson, MS, Area, 2006–2008
| Participants With HIV (n = 25), No. (%) | Participants Without HIV (n = 85), No. (%) | Adjusted Odds Ratio (95% Confidence Interval) | |
| Male sex partners older than 25 y during recall period | 15 (60) | 18 (21) | 5.5 (1.8, 17.3) |
| Unprotected anal intercourse with casual male sex partners during recall period | 12 (48) | 8 (9) | 6.3 (1.8, 22.3) |
| Would probably give in if a partner wanted to have unprotected sexa | 8 (32) | 6 (7) | 5.0 (1.2, 20.6) |
| Involved in a long-term relationship with or married to a mana | 9 (36) | 7 (8) | 5.3 (1.4, 20.7) |
Note. Adjusted odds ratios and confidence intervals were derived from multiple logistic regression analyses that included only the factors listed. The recall periods were the 12 months before the first positive HIV test for participants with HIV and the preceding 12 months for participants without HIV.
At the time of the first positive HIV test among participants with HIV or the time of the interview among participants without HIV.
For this logistic regression analysis, long-term relationship status and the factors listed in Table 3 (P ≤ .1) were variables eligible for inclusion. The use of forward, backward, and stepwise selection procedures yielded the same set of significant factors. When the selection process was repeated with age included as a continuous variable in addition to the 4 variables just described, the same set of factors was again significant, with little change in adjusted odds ratios. Therefore, because it had higher precision, the model described in Table 4 was the final model.
Not being in a long-term relationship with or married to a man and agreeing that one would give in to a partner who wanted to have unprotected sex interacted to significantly increase the odds of HIV infection. In a model that included the 4 variables listed in Table 4 as well as the interaction term, the odds of HIV infection among men who reported that they would give in to a partner who wanted unprotected sex varied depending on relationship status. Among those who were in a long-term relationship, reports of giving in were not significantly associated with HIV infection (adjusted OR = 0.1; 95% CI = 0.004, 2.7). However, among those who were not in a long-term relationship, such reports were associated with HIV infection (adjusted OR = 13.2; 95% CI = 2.6, 67.9).
DISCUSSION
Our analysis revealed that having an older male partner, engaging in unprotected anal intercourse with casual male partners, and being likely to give in if a partner wanted to have unprotected sex, especially when that partner was not a long-term one, were associated with HIV infection among young African American MSM in the Jackson, Mississippi, area.
Partner Age
We found that, among 16–25-year-old African American MSM, having a male sex partner older than 25 years was associated with HIV infection. This association was present for both younger (16–22 years old) and older (23–25 years old) participants (data not shown). Studies have shown that age-disassortative mixing (between older individuals who often have higher HIV prevalence and younger individuals who more frequently do not have HIV) probably contributes to the ongoing spread of HIV infection among heterosexual individuals in Africa11 and MSM in the United States.12–14 Specifically, the tendency toward having older partners combined with a high likelihood of having African American partners (both groups with higher HIV prevalence), as was the case among our participants, probably contributes to the high HIV acquisition among young African American MSM.12,15
It has been demonstrated that HIV prevalence rates are higher among older MSM.6,16 In our analysis, young African American MSM with partners from older age groups were not significantly more likely to report unprotected anal intercourse than those who did not report older partners (55% vs 52%; P = .8); however, we did not collect detailed data about unprotected sexual encounters, including the number of such encounters or whether they involved insertive unprotected anal intercourse or receptive unprotected anal intercourse, which carries a higher risk of HIV acquisition. Identifying determinants of partner selection with respect to age may reveal whether this behavior is modifiable and inform development of potential interventions.
Unprotected Anal Intercourse With Casual Male Partners
Our findings indicated that unprotected anal intercourse was associated with HIV infection. However, although unprotected anal intercourse with both main and casual partners was associated with HIV infection in the bivariate analysis, only unprotected anal intercourse with casual partners was an independent risk factor for HIV infection. A prior study identified anal sex with main partners as a risk factor for HIV infection among young MSM,2 and 2 studies have shown that sex with main partners probably accounts for the majority of HIV transmission among MSM,17 particularly young MSM.18 Because there is not a standard definition of main and casual partners, our definitions may have varied from those used in other studies. Moreover, men of different ages may interpret these terms and definitions differently, and such interpretations may affect reporting of risk behaviors.
Although unprotected anal intercourse with casual partners was the factor most strongly associated with HIV infection, the percentage of case participants and controls who reported unprotected anal intercourse with main partners (48%) was also higher than that found in other studies; 28% of African American MSM interviewed during 2004–2005 in a National HIV Behavioral Surveillance System survey reported unprotected anal intercourse with a main partner.19 Research is needed to understand determinants of unprotected anal intercourse among African American MSM in this region. Moreover, disseminating messages to African American MSM regarding the importance of engaging in protected sex with both main and casual partners may be important.
Self-Efficacy of Condom Use
Our results suggest that low condom use self-efficacy (a high likelihood of giving in to a partner's desire for unprotected sex) is associated with HIV infection. A high likelihood of giving in to a partner's desire for unprotected sex increased the odds of HIV infection among men who were not in a long-term relationship. One possible interpretation is that negotiating condom use with a long-term partner may differ from negotiating with a new or casual partner. This finding, together with the fact that the most common reason given for not using a condom was “we got caught up in the moment,” suggests a possible role for interventions that focus on improving sexual negotiation skills or for partnership counseling approaches.
Limitations
Our findings are subject to several limitations. All behaviors were self-reported. Whereas the recall period for controls was the preceding 12 months, the period for case participants was the 12 months before HIV diagnosis. In some instances, case participants were asked about behaviors that occurred up to 3 years before the interview. As a result, case participants may have been less likely to recall risk behaviors or may have recalled certain behaviors, such as number of partners, with less precision than controls.
In addition, our findings may be subject to selection bias because although we attempted to recruit all potential case participants, fewer than half were interviewed. After reviewing our data, we did not detect differences between those interviewed and those not interviewed with respect to demographic characteristics. However, those we were not able to interview may have been more transient and may have differed from those interviewed with respect to the behaviors studied. Case participants and controls who were excluded because they responded “don't know” or refused to answer questions about sexual behaviors or because they terminated the survey early may have been less willing to disclose the answers to these types of sensitive questions. This may reflect differences in openness regarding sexual behavior or may be an indicator of differences in behaviors relative to individuals included in the study.
The manner in which controls were recruited may also have resulted in selection bias. Those who were recruited at gay bars and clubs may have been more likely to self-identify as gay than those recruited elsewhere. Recent studies have shown that risk behavior levels with male sex partners are lower among African American MSM who do not self-identify as gay.20 We found that controls recruited at gay bars and clubs were somewhat more likely to self-identify as gay than those recruited elsewhere (73% vs 55%; P = .09) but were not more likely to engage in unprotected anal intercourse (50% vs 58%; P = .5).
In addition, recruitment at college campuses may have affected the likelihood of having partners from older age groups; however, those enrolled in college were not substantially less likely to report older partners than those not enrolled in college (28% vs 38%; P = .5). Controls recruited at an STD clinic may have been more likely than controls recruited elsewhere to engage in behaviors, such as unprotected anal intercourse, that would predispose them to contracting an STD. We recruited from several different venues in an attempt to minimize bias in any one direction. Young African American MSM in Mississippi are difficult to access. For this and other reasons, we are unable to assess whether our sample of controls is representative of young African American MSM in the Jackson area. Moreover, because the HIV status of controls was self-reported and the HIV test could have been administered as long as 6 months before the interview, misclassification of individuals with HIV as controls may have occurred.
Also, because this investigation was limited to the Jackson area, our results may not be generalizable to young African American MSM in other areas. Finally, the large burden of the HIV epidemic and other STDs in the South, and specifically the Deep South, is probably a result of a number of social, cultural, economic, and political factors, including lack of access to employment, medical care, and housing.21–24 Our investigation was not designed to address the role of these factors, which affect risk and prevention behaviors and access to and use of HIV testing and medical and social care among African Americans, including African American MSM.25,26
Conclusions
Our study has identified important HIV risk behaviors among young African American MSM with and without HIV in Mississippi, a state that has been the focus of few studies on this topic. Our findings provide several insights into the challenging and critical task of decreasing the number of new HIV infections among young African American MSM, especially in southern states that may have similar demographic characteristics, cultural dynamics, and socioeconomic challenges. Given the high prevalence of risk behaviors among the young African American MSM in our analysis, HIV prevention efforts must begin before or during early adolescence.
Our findings also support the idea that sexual negotiation and communication can play an important role in prevention. Further work is needed to understand the social, cultural, economic, and political factors that moderate risk behaviors and to develop prevention programs specifically designed to address the challenging yet critical task of reducing new HIV infections among young African American MSM.
Acknowledgments
This investigation was supported by the Centers for Disease Control and Prevention and the Mississippi State Department of Health.
We thank Kendra Johnson, Craig Thompson, Deborah Dowell, Anne McIntyre, Lisa Rynn, Teri Larkins, Jillian Doss, Michael Robinson, Anthony Fox, Mauda Monger, Linda Haynes, and Haitham Baghdady for their assistance with recruiting and interviewing participants for this investigation. We also thank Lisa Hightow-Weidman for assistance in preparation of the questionnaire, Greg Millett for input regarding study design and interpretation and his comments on an earlier version of this article, and Jeanne Bertolli for her comments on an earlier version of this article.
Human Participant Protection
This study was conducted in the context of a public health epidemiological investigation, and it was determined by the National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention at the Centers for Disease Control and Prevention (CDC) that the study did not require approval from the CDC or local institutional review boards.
References
- 1.Valleroy LA, MacKellar DA, Karon JM, et al. HIV prevalence and associated risks in young men who have sex with men. JAMA. 2000;284(2):198–204 [DOI] [PubMed] [Google Scholar]
- 2.Harawa NT, Greenland S, Bingham TA, et al. Associations of race/ethnicity with HIV prevalence and HIV-related behaviors among young men who have sex with men in 7 urban centers in the United States. J Acquir Immune Defic Syndr. 2004;35(5):526–536 [DOI] [PubMed] [Google Scholar]
- 3.HIV prevalence estimates—United States, 2006. MMWR Morb Mortal Wkly Rep. 2008;57(39):1073–1076 [PubMed] [Google Scholar]
- 4.Hall HI, Song R, Rhodes P, et al. Estimation of HIV incidence in the United States. JAMA. 2008;300(5):520–529 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Centers for Disease Control and Prevention Subpopulation estimates from the HIV Incidence Surveillance System—United States, 2006. MMWR Morb Mortal Wkly Rep. 2008;57(36):985–989 [PubMed] [Google Scholar]
- 6.Centers for Disease Control and Prevention Prevalence and awareness of HIV infection among men who have sex with men—21 cities, United States, 2008. MMWR Morb Mortal Wkly Rep. 2010;59(37):1201–1207 [PubMed] [Google Scholar]
- 7.Centers for Disease Control and Prevention Trends in HIV/AIDS diagnoses among men who have sex with men—33 states, 2001–2006. MMWR Morb Mortal Wkly Rep. 2008;57(25):681–686 [PubMed] [Google Scholar]
- 8.Millett GA, Flores SA, Peterson JL, Bakeman R. Explaining disparities in HIV infection among black and white men who have sex with men: a meta-analysis of HIV risk behaviors. AIDS. 2007;21(15):2083–2091 [DOI] [PubMed] [Google Scholar]
- 9.HIV transmission among black college student and non-student men who have sex with men—North Carolina, 2003. MMWR Morb Mortal Wkly Rep. 2004;53(32):731–734 [PubMed] [Google Scholar]
- 10.Centers for Disease Control and Prevention HIV infection among young black men who have sex with men—Jackson, Mississippi, 2006–2008. MMWR Morb Mortal Wkly Rep. 2009;58(4):77–81 [PubMed] [Google Scholar]
- 11.Gregson S, Nyamukapa CA, Garnett GP, et al. Sexual mixing patterns and sex-differentials in teenage exposure to HIV infection in rural Zimbabwe. Lancet. 2002;359(9321):1896–1903 [DOI] [PubMed] [Google Scholar]
- 12.Berry M, Raymond HF, McFarland W. Same race and older partner selection may explain higher HIV prevalence among black men who have sex with men. AIDS. 2007;21(17):2349–2350 [DOI] [PubMed] [Google Scholar]
- 13.Service SK, Blower SM. Linked HIV epidemics in San Francisco. J Acquir Immune Defic Syndr Hum Retrovirol. 1996;11(3):311–313 [DOI] [PubMed] [Google Scholar]
- 14.Service SK, Blower SM. HIV transmission in sexual networks: an empirical analysis. Proc R Soc Lond B Biol Sci. 1995;260(1359):237–244 [DOI] [PubMed] [Google Scholar]
- 15.Bingham TA, Harawa NT, Johnson DF, Secura GM, MacKellar DA, Valleroy LA. The effect of partner characteristics on HIV infection among African American men who have sex with men in the Young Men's Survey, Los Angeles, 1999–2000. AIDS Educ Prev. 2003;15(1, suppl A):39–52 [DOI] [PubMed] [Google Scholar]
- 16.MacKellar DA, Valleroy LA, Secura GM, et al. Unrecognized HIV infection, risk behaviors, and perceptions of risk among young men who have sex with men: opportunities for advancing HIV prevention in the third decade of HIV/AIDS. J Acquir Immune Defic Syndr. 2005;38(5):603–614 [DOI] [PubMed] [Google Scholar]
- 17.Sullivan PS, Salazar L, Buchbinder S, Sanchez TH. Estimating the proportion of HIV transmissions from main sex partners among men who have sex with men in five US cities. AIDS. 2009;23(9):1153–1162 [DOI] [PubMed] [Google Scholar]
- 18.Davidovich U, de Wit J, Albrecht N, Geskus R, Stroebe W, Coutinho R. Increase in the share of steady partners as a source of HIV infection: a 17-year study of seroconversion among gay men. AIDS. 2001;15(10):1303–1308 [DOI] [PubMed] [Google Scholar]
- 19.Sanchez T, Finlayson T, Drake A, et al. Human immunodeficiency virus (HIV) risk, prevention, and testing behaviors—United States, National HIV Behavioral Surveillance System: men who have sex with men, November 2003–April 2005. MMWR Surveill Summ. 2006;55(6):1–16 [PubMed] [Google Scholar]
- 20.Millett GA, Peterson JL, Wolitski RJ, Stall R. Greater risk for HIV infection of black men who have sex with men: a critical literature review. Am J Public Health. 2006;96(6):1007–1019 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Centers for Disease Control and Prevention Estimated number of AIDS cases among adult and adolescent men who have sex with men, by region and race/ethnicity, 2006—50 States and DC. Available at: http://www.cdc.gov/hiv/topics/surveillance/resources/slides/msm/slides/msm.pdf. Accessed September 17 2010
- 22.Doherty IA, Leone PA, Aral SO. Social determinants of HIV infection in the Deep South. Am J Public Health. 2007;97(3):391. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Reif S, Geonnotti KL, Whetten K. HIV infection and AIDS in the Deep South. Am J Public Health. 2006;96(6):970–973 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Farley TA. Sexually transmitted diseases in the southeastern United States: location, race, and social context. Sex Transm Dis. 2006;33(suppl 7):S58–S64 [DOI] [PubMed] [Google Scholar]
- 25.Peterson JL, Jones KT. HIV prevention for black men who have sex with men in the United States. Am J Public Health. 2009;99(6):976–980 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Friedman SR, Cooper HL, Osborne AH. Structural and social contexts of HIV risk among African Americans. Am J Public Health. 2009;99(6):1002–1008 [DOI] [PMC free article] [PubMed] [Google Scholar]